Extended differentiation of veins and stomata is essential for the expansion of large leaves in Rheum rhabarbarum

• 1967 Precise coordination between the demand for water at the

evapo-rating surfaces in the leaf and the supply of water to those surfaces appears to be a universal feature of plant leaves. This coordination is usually achieved by balancing stomatal density, which largely de-termines demand (Franks and Beerling, 2009), with leaf vein den-sity, which determines supply (Sack and Frole, 2006; Brodribb et al., 2007), through a mechanism that involves altering the densities of these two anatomical features through changes in cell size (Brodribb and Jordan, 2011; Carins Murphy et al., 2012). The developmental co-ordination between stomatal and vein density is driven by the selec-tive pressure to optimize the benefit- to- cost ratio of leaf construction such that an investment in a leaf vein network is sufficient to maintain stomata supplied with the minimum amount of water required to al-low them to open fully under saturating light conditions (Brodribb

and Jordan, 2011). In a number of developmental and hormone syn-thesis mutants, a loss of coordination of vein and stomatal density re-sults in reduced maximum leaf conductance and gas exchange (Dow and Bergmann, 2014; Dow et al., 2014; McAdam et al., 2017).

Evidence for the coordination between vein and stomatal den-sity has been found across and within species in response to vary-ing environmental conditions, such as light intensity and humidity (Brodribb and Jordan, 2011; Carins Murphy et al., 2012, 2016, 2017; Zhao et  al., 2017). A chief mechanism that induces these devel-opmental changes in vein and stomatal densities involves passive changes in cell size, especially epidermal pavement cells (Carins Murphy et al., 2012; Sack and Buckley, 2016; Carter et al., 2017). With all leaves initiating the same numbers of veins and stomata, when cell sizes change, the densities of veins and stomata change.

Extended differentiation of veins and stomata is essential

for the expansion of large leaves in

Rheum rhabarbarum

Amanda A. Cardoso1 _{, Joshua M. Randall}1_{, Gregory J. Jordan}2 _{, and Scott A. M. McAdam}1,3

Manuscript received 24 May 2018; revision accepted 31 August 2018.

1 _{Purdue Center for Plant Biology, Department of Botany and Plant}

Pathology, Purdue University, West Lafayette, IN 47907, USA

2 _{School of Natural Sciences, University of Tasmania, Hobart, TAS}

7001, Australia

3 _{Author for correspondence (e-mail: [email protected])}

Citation: Cardoso, A. A., J. M. Randall, G. J. Jordan, and S. A. M. McAdam. 2018. Extended differentiation of veins and stomata is essential for the expansion of large leaves in Rheum rhabarbarum. American Journal of Botany 105(12): 1967–1974.

doi:10.1002/ajb2.1196

PREMISE OF THE STUDY: The densities of veins and stomata govern leaf water supply and gas exchange. They are coordinated to avoid overproduction of either veins or stomata. In many species, where leaf area is greater at low light, this coordination is primarily achieved through differential cell expansion, resulting in lower stomatal and vein density in larger leaves. This mechanism would, however, create highly inefficient leaves in species in which leaf area is greater at high light. Here we investigate the role of cell expansion and differentiation as regulators of vein and stomatal density in Rheum rhabarbarum, which produces large leaves under high light.

METHODS: Rheum rhabarbarum plants were grown under full sunlight and 7% of full sunlight. Leaf area, stomatal density, and vein density were measured from leaves harvested at different intervals.

KEY RESULTS: Leaves of R. rhabarbarum expanded at high light were six times larger than leaves expanded at low light, yet vein and stomatal densities were similar. In high light- expanded leaves, minor veins were continuously initiated as the leaves expanded, while an extended period of stomatal initiation, compared to leaves expanded at low light, occurred early in leaf development.

CONCLUSIONS: We demonstrate that R. rhabarbarum adjusts the initiation of stomata and minor veins at high light, allowing for the production of larger leaves uncoupled from lower vein and stomatal densities. We also present evidence for an independent control of vein and stomatal initiation, suggesting that this adjustment must involve some unknown developmental mechanism.

KEY WORDS cell size; epidermis; leaf expansion; light intensity; stomatal development; stomatal density; vein density.

Low vein and stomatal densities can be found in large leaves in which cell expansion continues for a long period of development, whereas in leaves in which cell expansion is limited, leaf size is re-duced and vein and stomatal densities are higher (Carins Murphy et al., 2012). This passive determination of vein and stomatal density via variation in cell size is believed to be the primary mechanism by which leaves coordinate vein and stomatal densities in response to changes in environmental condition in most woody and herbaceous angiosperms (Carins Murphy et  al., 2012, 2016). Typically, leaves that expand under low light or low leaf- to- air vapor pressure dif-ference (VPD) have much larger cells and consequently lower vein and stomatal densities than leaves that expand under high light or high VPD (Carins Murphy et al., 2012, 2016). This reduction in vein and stomatal densities has a direct effect on the photosynthetic and hydraulic capacity of leaves (Sack and Frole, 2006; Brodribb et al., 2007; Xiong et al., 2014; Scoffoni et al., 2016; Zhang et al., 2018).

The passive dilution model appears to work well for a wide range of species, especially species that have larger leaves in the shade com-pared with sun. For those species, passive dilution can lead to a lower vein and stomatal density in the shade, without the need to change cell numbers. A reduced investment in veins and stomata in shade- adapted leaves is an optimal economical investment, given the lower photosyn-thetic yields of shade- grown leaves. However, many herbaceous species have a poor tolerance of shade and produce substantially larger leaves in the sun (Gay and Hurd, 1975; Dengler, 1980; Kalve et al., 2014). In such species, passive dilution without the addition of new veins and stomata would lead to lower vein and stomatal density under the very conditions where high densities, which are essential for high rates of gas exchange, are most beneficial.

As an alternative to passive cell expansion, the role of vein and stomatal differentiation during leaf expansion remains an understud-ied mechanism for developmental coordination of veins and stomata, particularly in species with larger leaves in the sun. The developmen-tal responses of veins and stomata in some fern species indicate that increased stomatal differentiation may indeed play a role in deter-mining the stomatal density of leaves grown under different light in-tensities (Carins Murphy et al., 2017). In addition, unlike major veins, which are formed at very early stages of leaf expansion, minor vein differentiation appears to be more flexible in terms of developmental timing, being continuously differentiated until the final stages of leaf expansion (Koizumi et al., 2000; Jun et al., 2002; Sack et al., 2012; Sack and Scoffoni, 2013). Thus, a prolonged initiation of minor veins could potentially lead to a maintenance of minor vein densities, regardless of leaf area or cell size, in large leaves (Sack et al., 2012).

Here we hypothesized that the coordination of vein and stoma-tal densities in a shade- intolerant species with large leaves in high light environments requires the active differentiation of veins and stomata during leaf expansion. To test this hypothesis, we analyzed leaf anatomical traits during leaf expansion in the geophyte species Rheum rhabarbarum L. (Polygonaceae), known for its capacity to produce very large leaves in high light environments and small leaves under low light.

MATERIALS AND METHODS

Plant material

Three potted individuals of R. rhabarbarum cv. Victoria were grown for the duration of the experiment at either high or low light intensity

from dormant, 5- mo- old rhizomes. Plants were grown in a controlled glasshouse under night/day temperatures of 16/22°C, respectively, and natural photoperiod of ca. 12 h. All plants were watered once per day and received liquid nutrients once per week. Plants grown at high light intensity received unobscured natural light, with a maximum photo-synthetic photon flux density (PPFD) of ca. 1500 μmol m–2 _s–1_{, while} plants grown at low light intensity were grown in the same environ-ment but under a light- reducing shade cloth, receiving a maximum PPFD of ca. 100 μmol m–2 _s–1_{. Excluding the first three leaves} follow-ing the breakfollow-ing of rhizome dormancy, which may have been initiated under different environmental conditions, all subsequent leaves were tagged on the day of first emergence from ochrea. Plants grown in high light produced the same number of leaves as those in low light during the experimental period. After 1 month of growth, with all emerged leaves tagged, leaves were harvested so that the collection contained leaves that spanned the full developmental range from initial emer-gence from ochrea to fully expanded leaves under each light condition. All leaves were immediately stored at −20° C until later processing for anatomy. The age range was defined based on a preliminary experi-ment carried out with R. rhabarbarum, which demonstrated that leaves take 4 weeks to fully expand under both light conditions (Appendix S1, see Supplemental Data with this article). The first day of leaf expansion was considered to be the first day of visible leaf emergence from the ochrea (Appendix S2), but it should be noted that vein and stomatal initiation began before that stage.

Leaf area and anatomical traits during leaf expansion

Bethesda, MD, USA). Finally, total minor vein length and total stomatal and epidermal cell number were calculated by multi-plying their density by leaf area for each leaf, which was scanned after excising through undulations to flatten completely with no leaf overlap on a flatbed scanner.

Dilution models during leaf expansion

We compared modelled relationships that assume vein and stomatal density dilution occurred due to passive increases in cell size during leaf expansion starting at day 5 of leaf emergence from the ochrea with the observed relationships of vein and sto-matal densities under different light conditions. Day 5 was cho-sen as the starting point for the model to allow for a period of cell differentiation before passive expansion. The leaf growth rate for each light condition was used to determine the rate of dilution during expansion, using initial leaf area and either initial vein or stomatal density, considering no new vein or stomatal differenti-ation. The mean vein and stomatal dilution based on passive cell expansion were calculated using leaf growth rate at either high or low light intensity and initial values for mean leaf area, total vein length or total stomatal number, respectively. For that, the following equations were used:

and

Initial leaf area, total vein length, and total stomatal number were assumed to be equivalent to the mean measurements of leaves at day 5, as we assumed 5 d was sufficient time to initiate all veins and stomata after which passive dilution occurred. Leaf growth rate was also obtained using data of leaves from 5 to 28 d following emer-gence from the ochrea.

Major vein density of fully expanded leaves at low and high light

The density of major veins (i.e., first- to fourth- order veins) was analyzed from images of intact leaves. Vein diameter was not used to categorize vein orders, as diameters from the same vein orders varied between sun and shade leaves, instead we categorized vein orders by branching pattern (Sack et  al., 2012; Wen et  al., 2018). Leaves of R. rhabarbarum are palmately veined (Niinemets et al., 2007), with around five first- order veins branching from the pet-iole. The second- order veins were defined as those that branched from the first- order veins, and the third- and fourth- order veins were defined as those branched from the second- and third- order veins, respectively (Hickey, 1973). The density of major veins was accessed for three fully expanded leaves (28 d following emergence from ochrea) formed at either low or high light.

Statistical analyses

Data were collected from three plants grown at high light and three grown at low light. The nonlinear associations between the

morphophysiological features were analyzed by generalized addi-tive models using the gam function in the mgcv package (Wood and Wood, 2015) in R (R Core Team, 2014), assuming a normal distribution and with the recommended default global value chains smoothing. The significance of the models was tested by ANOVA, and similarities between the slopes for the two light conditions were tested by ANCOVA. The linear relationship between the major vein density and leaf area was also tested by ANOVA.

RESULTS

The increase in leaf area in leaves expanded in high light was 6- fold greater than leaves expanded at low light (Fig. 1). Leaves of R. rhabarbarum underwent a rapid expansion phase until the 20th day following emergence from ochrea when leaf expansion ceased

Minor vein density at dayt= Initial total minor vein length

Initial leaf area+₍t ×_{Leaf growth rate)}

Stomatal density at dayt= Initial total stomatal number

Initial leaf area+₍t ×_{Leaf growth rate)}.

FIGURE 1. (A) Leaf area during leaf expansion in Rheum rhabarbarum

plants grown under either high (red symbols) or low light intensity (blue symbols). Generalized additive model curves; *P < 0.05; **P < 0.01; ***P

< 0.001, the banding on either side of the model represents the 95% confidence interval. (B) Representative image of 28- d- old leaves fully ex-panded under either high or low light. Day 1 corresponds to the first day a leaf visibly emerged from ochrea.

A

0

Leaf area (cm

2 )

Leaf age (days)

0 5 10 15 20 25

10 cm

High light

Low light

B

800

600

400

abruptly (Fig. 1; P < 0.001; Appendix S1). Despite considerable dif-ferences in leaf area between high- and low- light- expanding leaves, minor vein density and stomatal density showed similar patterns of decline with age, so that, at any age leaves expanding under both light conditions had similar values (Fig.  2A, B). The patterns for stomatal density and minor vein density with respect to age were subtly different (P < 0.001), with a continuous decline in stomatal density starting at the 5th day (Fig.  2B), but a more rapid initial decline (over the first 10 d following leaf emergence from ochrea) in minor vein density (Fig.  2A). The relationship between minor vein density and stomatal density in leaves expanding under low light was slightly steeper from that in high- light leaves (P < 0.001; Fig. 2C). Although both low- and high- light- expanding leaves had a higher number of stomata per minor vein length at the begin-ning compared to the end of leaf expansion, high- light- expanding leaves had more stomata per unit of minor vein than did low- light- expanding leaves at full expansion (Fig. 2C).

Epidermal cell size increased in a similar way during the expansion of both high- and low- light- expanding leaves, with a single regression explaining epidermal cell growth under both light conditions (Fig. 3). For both high- and low- light- expanding leaves, minor vein density declined in response to initial increases in the epidermal cell size, and then it was maintained despite continuous increases in epidermal cell size (P < 0.001; Fig. 3). Stomatal den-sity, however, was high and changed little as leaves began to expand, but declined rapidly when epidermal cell size measurably increased (P < 0.001; Fig. 3).

The total number of epidermal cells was constant during leaf expansion in leaves grown under both high and low light inten-sities, with little differentiation of new cells as leaves expanded (P > 0.05; Fig. 4A). In contrast, total minor vein length increased as leaves continuously expanded in both light conditions, yet at different rates (P < 0.01 for high light and P < 0.05 for low light; Fig.  4B). While low- light- expanded leaves maintained a con-stant number of stomata during leaf expansion, the total num-ber of stomata increased considerably in high- light- developed leaves (P < 0.05 for high light and P > 0.05 for low light; Fig. 4C). During the first 10 d, a higher density of stomatal primordia was found in leaves grown in high light compared with leaves grown in low light (Appendix S4). Stomatal primordia in high- light- developed leaves diminished in density until the 10th day of leaf emergence from ochrea and after that were no longer observed (Appendices S3, S4).

Observed data for vein and stomatal density during leaf ex-pansion at low light was very similar to the predicted relation-ships between either vein or stomatal density and leaf age, which assumed dilution in density solely due to increases in leaf area and not differentiation (Fig.  5). However, observed minor vein density was slightly higher than the modelled minor vein density through time, suggesting some vein development as leaves ex-panded in low light. In high- light- exex-panded leaves, the observed data of vein and stomatal density changes as leaves expand could not be explained by passive dilution through epidermal cell ex-pansion alone, implying that new veins and stomata were initiated as leaves expanded at high light intensity (Fig. 5). The density of major veins was highly correlated with leaf area for fully expanded leaves grown at both light conditions, large high- light- expanded leaves had a lower major- vein density (0.49 ± 0.01) compared to the smaller leaves that expanded under low light (0.107 ± 0.04) (Appendix S5).

FIGURE 2. (A) Minor vein and (B) stomatal density during leaf expansion and (C) relationship between minor vein density and stomatal density during leaf expansion in Rheum rhabarbarum plants grown under either high (red symbols) or low light intensity (blue symbols). In (C) different color shades represent different ranges of leaf age. Generalized additive model curves: test of significant relationship between the variables: *P

< 0.05; **P < 0.01; ***P < 0.001, the banding on either side of the model represents the 95% confidence interval. Day 1 corresponds to the first day a leaf visibly emerged from ochrea.

Stomatal density (m

m

-2 )

r2 _{= 0.66***}

r2 _{= 0.66***}

50 0

2 4 6 8 10 12 0 2 4 6 8 10 12

Stomatal density (mm-2₎

Minor vein density (mm mm

-2 ) r

2 _{= 0.59***}

r2 _{= 0.86***}

A

B

C

100 150 200

Leaf age (days) Leaf age (days)

Minor vein density (mm mm

DISCUSSION

This study demonstrates that in a herbaceous species with a large degree of plasticity in leaf area, vein and stomatal densities are coordinated by the regulation of minor vein and stomatal

differentiation during leaf expansion. This regulatory mechanism contrasts with many other herbaceous and woody plants that regu-late stomatal and vein densities passively by changing cell size and not via differentiation (Carins Murphy et al., 2012, 2016). The ad-dition of extra veins and stomata provides a mechanism that allows the shade- intolerant geophyte, R. rhabarbarum, to achieve large leaf area with high vein and stomatal densities in high- light en-vironments. This mechanism may be especially important for the many herbaceous geophyte species in which leaves are the dom-inant aboveground part of the plant, and the expansion of large leaves minimizes competition with other shade- intolerant compet-itors. In species where a strong selective pressure is imposed for large leaves in high- light environments, the active differentiation of veins and stomata would allow plants to maximize leaf area with-out decreasing leaf gas- exchange capacity or increasing suscepti-bility to damage by desiccation.

Continuous development of minor veins and prolonged initiation of stomata during ongoing leaf expansion at high light

Major vein density was highly correlated with leaf area in both high- and low- light- expanded leaves, decreasing as leaf area increased (Appendix S5). Declines in major vein density with increasing leaf area are expected given that these veins are formed in limited quantities in the primordial leaf (Sack et al., 2012), and for this reason, final major vein density exhibits a strong correla-tion with leaf area (Sack et al., 2012). Unlike declines in major vein density driven by increases in leaf area, the same stomatal and mi-nor vein densities were observed in low- and high- light- expanded leaves of R. rhabarbarum despite 6- fold differences in leaf area (Fig. 2A). This similarity in vein density across leaf sizes in this species can be explained by the active differentiation of new minor veins and stomata during leaf expansion in large leaves grown un-der high light (Fig. 4). Light quantity has been long demonstrated to induce stomatal development (Schoch et  al., 1980; Lee et  al., 2017), and here we also observe a prolonged initiation of veins under higher light intensity.

The window for active differentiation for minor veins and sto-mata at high- light intensity seemed to be subtly different. Vein differentiation occurred over the whole period of leaf expansion (Fig.  4B). This continuous development of minor veins during leaf expansion was predicted in Arabidopsis thaliana (Sack et al., 2012). In contrast, the initiation of large numbers of new stomata (and epidermal cells) occurred only until the 10th day following leaf emergence from ochrea, corresponding with the age of leaves in which stomatal primordia were present and leaves reached a max-imum rate of expansion (Appendix S4). A short period of stomatal initiation has been observed in the small- leaved- species Arabidopsis thaliana (Nadeau and Sack, 2002). The continuous development of FIGURE 3. (A) Epidermal cell size during leaf expansion and (B) the re-lationship between minor vein and (C) stomatal density with epidermal cell size in Rheum rhabarbarum plants grown under either high (red sym-bols) or low light intensity (blue symsym-bols). Generalized additive model curves: test of significant relationship between the variables: *P < 0.05; **P < 0.01; ***P < 0.001, the banding on either side of the model repre-sents the 95% confidence interval. Day 1 corresponds to the first day a leaf visibly emerged from ochrea.

10 12

Epidermal cell size (m

m

2 x 10 -3)

Minor vein density (mm mm

-2)

50 100 150 200 250

Stomatal density (m

m

-2)

Epidermal cell size (mm2 _{x 10}-3₎

Epidermal cell size (mm2 _{x 10}-3₎

A

B

C

r2 _{= 0.70***}

r2 _{= 0.67***}

r2 _{= 0.80***}

r2 _{= 0.82***}

10

6 8 12

0 5 10 15 20 25

Leaf age (days) 0

2 4 6 8 10 12

0 2 4 6 8 10 12

minor veins yet shorter period of stomatal initiation suggests that final leaf area may be developmentally determined from an early stage of leaf expansion and closely related to the degree of stomatal initiation during the first 10 days of leaf expansion.

Vein and stomatal densities are controlled by independent mechanisms

Our observations in R. rhabarbarum add to a growing body of ev-idence suggesting that highly independent regulatory mechanisms control the development of veins and stomata, despite the strong de-velopmental coordination in the densities of these two key anatom-ical features (Carins Murphy et al., 2017; McAdam et al., 2017). The initiation of veins and stomata has been extensively described yet no apparent cross- talk between these two coordinated features of plant leaves has been suggested. The development of a leaf vein network has been linked to auxin biosynthesis and signaling, with auxin- mutants presenting reduced vein densities (Tobeña- Santamaria et al., 2002; Verna et al., 2015; McAdam et al., 2017). These mutations have been observed to result in considerable impaired vein forma-tion independently of any changes in stomatal density or anatomy (McAdam et al., 2017). Stomatal development, involves several well- described stages, including asymmetric and symmetric divisions from meristemoids into a pair of guard cells (Pillitteri et al., 2007). These stages have been demonstrated to be spatially and tempo-rally regulated by the key genetic regulators SPEECHLESS (SPCH), MUTE, and FAMA (Pillitteri et al., 2007) as well as inducer of CBF expression (ICE) transcription factors (Kanaoka et al., 2008). Like angiosperms (Schoch et al., 1980; Lee et al., 2017), fern stomatal initi-ation has been found to be augmented by high light intensity (Carins Murphy et al., 2017) and, interestingly, this increased stomatal differ-entiation at high light in ferns was independent of modifications in the leaf vein network (Carins Murphy et al., 2017).

The temporally independent differentiation of veins and sto-mata, allows the deviation in the proportional relationship between the densities of these two features to occur during leaf expansion (Fig. 2C). The higher production of veins than stomata during the final stages of leaf expansion in R. rhabarbarum resulted in the highest vein length per number of stomata being observed when leaves had fully expanded. The lower vein to stomatal ratio in young, incompletely expanded leaves of this species suggests undersupply of water to the stomata in these leaves places a considerable hydrau-lic limitation on stomatal opening.

CONCLUSIONS

We demonstrate here that, unlike most shade- tolerant species, the shade- intolerant species Rheum rhabarbarum can adjust the initia-tion of stomata and minor veins under high- light environments, fa-cilitating the production of larger leaves uncoupled from lower vein and stomatal densities. Such a mechanism is likely to be a critical determinant of leaf anatomy in a wide range of shade- intolerant geo-phytes, in which large leaves provide a strong competitive advantage in high- light environments. Given strong evidence for independent FIGURE 4. (A) Total number of epidermal cells, (B) minor vein length, and (C) number of stomata in leaves of Rheum rhabarbarum plants dur-ing leaf expansion grown under either high (red symbols) or low light intensity (blue symbols). Generalized additive model curves: test of sig-nificant relationship between the variables: *P < 0.05; **P < 0.01; ***P < 0.001, the banding on either side of the model represents the 95% con-fidence interval. Day 1 corresponds to the first day a leaf visibly emerged from ochrea.

r2 _{= 0.53**}

r2 _{= 0.10*}

B

0 50 100 150

0 100 200 300 400 500

0 5 10 15 20 25

Leaf age (days) Leaf age (days)

To

tal minor vein length (mm x 10

4)

A

C

To

tal number of epidermal cells (x 10

5)

To

tal number of stomata (x 10

4) r

2 _{= 0.31*}

Leaf age (days)

0 5 10 15 20

0 5 10 15 20 25

25

controls on the initiation of veins and stomata, this adjustment must involve some form of developmental control that is both currently unknown and which is not active in shade- tolerant species.

ACKNOWLEDGEMENTS

We gratefully acknowledge the advice of Tim Brodribb, Madeline Carins Murphy, and two anonymous reviewers and the editor. We also thank Justine Krueger for performing preliminary observations and Krista Johnson for helping cultivate plants and collect the leaves.

AUTHOR CONTRIBUTIONS

S.A.M.M. and G.J.J. designed the study; A.A.C. and J.M.R. carried out the experiments; A.A.C., S.A.M.M. and G.J.J. analyzed the data; A.A.C., S.A.M.M. and G.J.J. wrote and revised the manuscript.

SUPPORTING INFORMATION

Additional Supporting Information may be found online in the supporting information tab for this article.

FIGURE 5. Modelled (lines) and observed (symbols) (A, C) minor vein density and (B, D) stomatal density during leaf expansion in Rheum rhabarbarum

plants grown under either high (red symbols) or low light intensity (blue symbols). Observed data are taken from Fig 2. Modelled data, presented as the black lines, assume vein and stomatal density dilution due to increases in leaf area in the absence of new vein and stomatal differentiation during leaf expansion (see the Materials and Methods for details).

Leaf age (days)

Leaf age (days)

Stomatal density (m

m

-2

)

Stomatal density (m

m

-2

)

High light

Low light

A

B

C

LITERATURE CITED

Brodribb, T. J., and G. J. Jordan. 2011. Water supply and demand remain bal-anced during leaf acclimation of Nothofagus cunninghamii trees. New Phytologist 192: 437–448.

Brodribb, T. J., T. S. Feild, and G. J. Jordan. 2007. Leaf maximum photosynthetic rate and venation are linked by hydraulics. Plant Physiology 144: 1890–1898. Carins Murphy, M. R., G. J. Jordan, and T. J. Brodribb. 2012. Differential leaf

expansion can enable hydraulic acclimation to sun and shade. Plant, Cell & Environment 35: 1407–1418.

Carins Murphy, M. R., G. J. Jordan, and T. J. Brodribb. 2016. Cell expansion not cell differentiation predominantly co- ordinates veins and stomata within and among herbs and woody angiosperms grown under sun and shade.

Annals of Botany 118: 1127–1138.

Carins Murphy, M. R., G. J. Jordan, and T. J. Brodribb. 2017. Ferns are less de-pendent on passive dilution by cell expansion to coordinate leaf vein and stomatal spacing than angiosperms. PLOS One 12: e0185648.

Carter, R., Y. E. Sánchez-Corrales, M. Hartley, V. A. Grieneisen, and A. F. Marée. 2017. Pavement cells and the topology puzzle. Development 144: 4386–4397. Dengler, N. G. 1980. Comparative histological basis of sun and shade leaf

dimorphism in Helianthus annuus. Canadian Journal of Botany 58: 717–730.

Dow, G. J., and D. C. Bergmann. 2014. Patterning and processes: how stoma-tal development defines physiological potential. Current Opinion in Plant Biology 21: 67–74.

Dow, G. J., J. A. Berry, and D. C. Bergmann. 2014. The physiological impor-tance of developmental mechanisms that enforce proper stomatal spacing in

Arabidopsis thaliana. New Phytologist 201: 1205–1217.

Fiorin, L., T. J. Brodribb, and T. Anfodillo. 2016. Transport efficiency through uniformity: organization of veins and stomata in angiosperm leaves. New Phytologist 209: 216–227.

Franks, P. J., and D. J. Beerling. 2009. Maximum leaf conductance driven by CO₂ effects on stomatal size and density over geologic time. Proceedings of the National Academy of Sciences, USA 106: 10343–10347.

Gay, A. P., and R. G. Hurd. 1975. The influence of light on stomatal density in the tomato. New Phytologist 75: 37–46.

Hickey, L. J. 1973. Classification of architecture of dicotyledonous leaves.

American Journal of Botany 60: 17–33.

Jun, J. H., C. M. Ha, and H. G. Nam. 2002. Involvement of the VEP1 gene in vas-cular strand development in Arabidopsis thaliana. Plant and Cell Physiology

43: 323–330.

Kalve, S., J. Fotschki, T. Beeckman, K. Vissenberg, and G. T. S. Beemster. 2014. Three- dimensional patterns of cell division and expansion throughout the development of Arabidopsis thaliana leaves. Journal of Experimental Botany

65: 6385–6397.

Kanaoka, M. M., L. J. Pillitteri, H. Fujii, Y. Yoshida, N. L. Bogenschutz, J. Takabayashi, J. K. Zhu, and K. U. Torii. 2008. SCREAM/ICE1 and SCREAM2

specify three cell- state transitional steps leading to Arabidopsis stomatal dif-ferentiation. Plant Cell 20: 1775–1785.

Koizumi, K., M. Sugiyama, and H. Fukuda. 2000. A series of novel mutants of

Arabidopsis thaliana that are defective in the formation of continuous vascu-lar network: calling the auxin signal flow canalization hypothesis into ques-tion. Development 127: 3197–3204.

Lee, J. H., J. H. Jung, and C. M. Park. 2017. Light inhibits COP1- mediated deg-radation of ICE transcription factors to induce stomatal development in Arabidopsis. Plant Cell 29: 2817–2830.

McAdam, S. A. M., M. P. Eléouët, M. Best, T. J. Brodribb, M. C. Murphy, S. D. Cook, M. Dalmais, et al. 2017. Linking auxin with photosynthetic rate via leaf venation. Plant Physiology 175: 351–360.

Nadeau, J. A., and F. D. Sack. 2002. Control of stomatal distribution on the

Arabidopsis leaf surface. Science 296: 1697–1700.

Niinemets, Ü., A. Portsmuth, and M. Tobias. 2007. Leaf shape and venation pattern alter the support investments within leaf lamina in temperate species: a neglected source of leaf physiological differentiation? Functional Ecology 21: 28–40. Pillitteri, L. J., D. B. Sloan, N. L. Bogenschutz, and K. U. Torii. 2007. Termination of

asymmetric cell division and differentiation of stomata. Nature 445: 501–505. R Core Team. 2014. R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria.

Sack, L., and T. N. Buckley. 2016. The developmental basis of stomatal density and flux. Plant Physiology 171: 2358–2363.

Sack, L., and K. Frole. 2006. Leaf structural diversity is related to hydraulic ca-pacity in tropical rain forest trees. Ecology 87: 483–491.

Sack, L., and C. Scoffoni. 2013. Leaf venation: structure, function, development, evolution, ecology and applications in the past, present and future. New Phytologist 198: 983–1000.

Sack, L., C. Scoffoni, A. D. McKown, K. Frole, M. Rawls, J. C. Havran, H. Tran, and T. Tran. 2012. Developmentally based scaling of leaf venation architec-ture explains global ecological patterns. Nature Communications 3: 837. Salisbury, E. J. 1927. On the causes and ecological significance of stomatal

frequency, with special reference to the woodland flora. Philosophical Transactions of the Royal Society, B, Biological Sciences 216: 1–65.

Schoch, P. G., C. Zinsou, and M. Sibi. 1980. Dependence of the stomatal index on environmental factors during stomatal differentiation in leaves of Vigna sinen-sis L.: 1. Effect of light intensity. Journal of Experimental Botany 31: 1211–1216. Scoffoni, C., D. S. Chatelet, J. Pasquet-kok, M. Rawls, M. J. Donoghue, E. J.

Edwards, and L. Sack. 2016. Hydraulic basis for the evolution of photosyn-thetic productivity. Nature Plants 2: 16072.

Tobeña-Santamaria, R., M. Bliek, K. Ljung, G. Sandberg, J. N. M. Mol, E. Souer, and R. Koes. 2002. FLOOZY of petunia is a flavin mono- oxygenase- like pro-tein required for the specification of leaf and flower architecture. Genes & Development 16: 753–763.

Verna, C., M. G. Sawchuk, N. M. Linh, and E. Scarpella. 2015. Control of vein network topology by auxin transport. BMC Biology 13: 94.

Wen, W., B. Li, B. Li, and X. Guo. 2018. A leaf modeling and multi- scale remesh-ing method for visual computation via hierarchical parametric vein and margin representation. Frontiers in Plant Science 9: 783.

Wood, S., and M. S. Wood. 2015. Package ‘mgcv’. R package version 1, 29. Xiong, D., T. Yu, T. Zhang, Y. Li, S. Peng, and J. Huang. 2014. Leaf hydraulic

conductance is coordinated with leaf morpho- anatomical traits and nitrogen status in the genus Oryza. Journal of Experimental Botany 66: 741–748. Zhang, F. P., M. R. Carins Murphy, A. A. Cardoso, G. J. Jordan, and T. J. Brodribb.

2018. Similar geometric rules govern the distribution of veins and stomata in petals, sepals and leaves. New Phytologist 219: 1224–1234.